Extensive studies of the Carcinoembryonic Antigen of the Human Digestive System (CEA) have failed to establish that this glycoprotein is either tumor specific or confined to entodermal structures. Nor has the relation of CEA production to tumor behavior been elucidated completely. It is also not known if the production of abnormal quantities or types of glycoprotein is a phenomenon confined to human colonic cancers or a more widespread feature of these tumors in other species, with perhaps a fundamental role in tumor growth and spread. Efforts directed at explaining the relationship of CEA or other glycoproteins to the malignant process require prospective studies of the time of appearance, location and production of CEA and like substances. Studies of the glycoproteins of existing defined animal models for large bowel cancer, therefore, appear worthwhile. Tumors will be developed in syngeneic rats using N-Methyl-N'-Nitro- N-Nitro- sequanidin (MNNG) and dimethylhydrazine (DMH) as described by others. Techniques already developed in this laboratory will be used to determine if such cancers produce abnormal glycoproteins analogous to CEA and to establish the time of their appearance in the process of neoplastic transformation and their relation to the biologic behavior of the tumors. If such antigens are detected, later studies will examine their structure and possible role in evoking host response or in protecting the tumor cells from host defenses. In addition to providing an understanding of CEA production, it is possible that these studies may allow the development of an assay for serum "CEA" in the rat which may be of value in following the effects of therapeutic measures in experimental models of large bowel cancer.